near term ToDos

[kjappelbaum]

Code

• Benchmark bencher (it seemed slow)
• Implement PBE bencher
• release mol-tda in pypi #2 -> contacted Aditi about PR
• enhance conda recipe #202
• provide the same set of pre-computed features for datasets to allow consistent use of the splitting strategies?  #239 -> running
• publish moffragmentor #173
• add docstrings for bencher #264
• add some installation instructions while the conda things are still waiting to be merged #265
• do we add a flavor parameter to the qmof dataset that will be there while we still compute the missing GCMC data? #268
• add note in docs about temporary directories #269
• consistent column names in dataset dataframes #271
• in splitters, automatically adjust number of bins if users do not specify #274
• add note about OS compatability to README #278
• Include WSL and macOS in GH actions #279
• add note about assumptions for default de-duplication  #282
• hyperopt tutorial
• update all datasets with features + hashes #285
• Run benchmark rebuild upon PR #222 Update Leaderboards #326
• Upload AiiDA DB to MaterialsCloud -> verdi export ran out of memory ...
• dataset to mpcontrib -> depends on MC upload to get the DOI (Hosting QMOF Data on MPContribs #318 )
• Add scripts for creating datasets to dev_scripts #360
• Complete documentation for labels in dataset #359
• QMOF pre-processing already drops basename duplicates. #322
• GCMC column labels for QMOF are not meaningful #354
• move datasets to Zenodo #361
• Add densities to ARC-MOF datasset #362

Paper

• New splitter figure
• Re-run duplicate experiment -> running (with core and bw)
• Run-run splitter experiment -> running
• Make sure we cite the model evaluation review

Ask for feedback

• Fergus
• Philippe

[kjappelbaum]

closing this meta issue for cleanup